Data managed by a storage system may be used in various applications such as search and analysis.
For example, in big-data analysis, analysis of unstructured data in which a storage structure of files and the like is yet to be determined is anticipated as a potentially useful method of obtaining new findings and realizations in business. In big-data analysis, in order to prevent situations where searches take time due to analyzing a large amount of data and completing the analysis becomes a much time-consuming process, a set made up of only data necessary for analysis may be created from the large amount of data. A set made up of only necessary data is referred to as a “data mart” (hereinafter, DM) and creating the data set is referred to as a “DM creation process”.
Generally, in big-data analysis, a DM creation process takes a large amount of time. This is because a process of duplicating (extracting) data necessary for data analysis from a large amount of data and storing the duplicated (extracted) data in the DM is time consuming. PTL 1 discloses a technique that enables data to be duplicated in a short period of time by presenting source data to a host computer as virtually-duplicated snapshot data.